BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Gaussian and non-Gaussian universality\, with applications to data
  augmentation - Peter Orbanz (UCL)
DTSTART:20250214T140000Z
DTEND:20250214T150000Z
UID:TALK226144@talks.cam.ac.uk
CONTACT:Qingyuan Zhao
DESCRIPTION:The term Gaussian universality refers to a class of results th
 at are\, loosely speaking\, generalized central limit theorems (where\, so
 mewhat confusingly\, the limit law is not necessarily Gaussian). They prov
 ide useful tools to study certain problems in machine learning. I will giv
 e a short overview of this idea and then present two types of results: One
  are upper and lower bounds that map out where Gaussian universality is ap
 plicable and what rates of convergence one can expect. The other is the us
 e of these techniques to obtain quantitative results on the effects of dat
 a augmentation in machine learning problems.\n
LOCATION:Centre for Mathematical Sciences MR15\, CMS
END:VEVENT
END:VCALENDAR